Novel regulatory SNPs that can be activated due to metformin treatment may orchestrate liver gluconeogenesis and add to the variability in AMPK-dependent mechanisms of metformin response

Metformin is a first-line therapy for type 2 diabetes, yet individual response varies significantly, with over 30 % of patients failing to achieve optimal glycemic control. The specific regulatory mechanisms of this phenomenon remain poorly understood and genetic variants involved are mainly undiscovered. There are multiple lines of evidence that the leading role in determining the variance in phenotypic outcome belongs to regulatory SNPs (rSNPs) as they directly modify gene expression. Therefore, the genome-wide search for such functional variants and deciphering associated phenotypes stands as a fundamental challenge. Previously, based on the results of bioinformatics analysis of allele-specific expression and binding landscape in human peripheral blood mononuclear cells, we have established an original panel of 14 796 rSNPs within promotors of 5132 genes. Aiming to pinpoint functional variants most likely linked to metformin hepatic response and impacts on liver gluconeogenesis, we analyzed the relevant open-access data as well as rSNPs from our panel and the corresponding genes. 1196 genes reported to be regulated by metformin in human hepatocytes and 115 genes involved in gluconeogenesis and/or its regulation via Gene Ontology annotations were intersected. Free R software and STRING v.11 tools were used for functional annotation. A number of genes harboring rSNPs within promotor regions were found to be particularly implicated in the mechanisms of metformin's action. Functional enrichment analyses revealed enrichment in critical pathways including FoxO, TNF-α and TGF-β signaling, also implicated in diabetes complications. Among these, six genes (ARPP19, ATF4, NR3C1, PFKFB3, TCF7L2, and WDR5) were strongly associated with regulation of gluconeogenesis, and may be modulated by metformin in the liver. We conclude that metformin therapy response may be influenced by the newly identified functional SNPs including rSNPs within the promotors of genes for gluconeogenic enzymes and transcription regulators.